Fenofibrate in Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Fenofibrate in Type 2 Diabetes (NCT03829514), a na phase 4 study. The primary outcome is Number of Individual Plasma Proteins That Changed From Baseline to End-point Based on Limma T-Test of Protein Abundance as Determined by Proteomic Analysis Via Liquid Chromatography-mass Spectrometry.

2. Clinical Question

The registry frames the study as an investigation of novel biomarkers and mechanisms related to fenofibrate in type 2 diabetes. Its summary describes interest in understanding how the drug may affect vulnerable tissues and in identifying possible targets for future research. Those aims provide context, but the listed primary outcome is the count of proteins meeting the analysis criterion for change. It does not report a clinical outcome measuring diabetic complications or tissue protection.

Because the study has one arm, any baseline-to-endpoint pattern is a within-participant observation over time. Without a separate control group, the design cannot separate changes associated with fenofibrate from changes due to time, other influences, or measurement variability. The record does not report a comparative treatment effect.

3. Trial Design

This is a completed Phase 4, single-group study of fenofibrate in type 2 diabetes, registered as NCT03829514. The sponsor is the Medical University of South Carolina. The record lists enrollment as 10 and describes one arm in which participants took the study medication. Allocation is listed as NA, and masking is listed as none. The record therefore does not describe a randomized, masked comparison with a control arm.

The stated study period began on 2019-02-04, and primary completion was recorded as 2020-03-04. The primary endpoint covers six weeks, from baseline visit to study completion visit. Those dates describe the study record; the endpoint timeframe is the period over which the paired protein measurements were compared.

The design is suited to describing within-participant protein measurements over the stated period. In general, comparing a participant with their own baseline can account for some stable differences between people. It does not, by itself, provide the counterfactual comparison needed to determine what would have happened over the same period without the intervention. As the registry explicitly notes, the arms do not form a like-for-like control comparison.

The record does not report a sample-size justification, protocol details beyond the listed information, interim analyses, or a full statistical analysis plan. It also does not describe additional study arms or a randomized assignment process. The posted endpoint analysis is based on protein abundance measured through proteomic analysis by liquid chromatography-mass spectrometry. A participant sample was unavailable for testing because it lacked quantity, as noted in the endpoint population description.

The registry status is completed. This design and status establish the setting for interpreting the result, but they do not establish that an observed within-participant change was caused by fenofibrate. The primary result should be read as a report about the proteins meeting the stated statistical criterion in this study, rather than as a controlled estimate of treatment effect.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: One participant sample lacked quantity to test.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Number of Individual Plasma Proteins That Changed From Baseline to End-point Based on Limma T-Test of Protein Abundance as Determined by Proteomic Analysis Via Liquid Chromatography-mass Spectrometry, measured over Six weeks, from baseline visit to study completion visit.

The primary endpoint is the number of individual plasma proteins that changed from baseline to the endpoint according to a Limma t-test of protein abundance. Protein abundance was determined from intensity values measured by proteomic analysis using liquid chromatography-mass spectrometry. The endpoint timeframe is six weeks, from baseline visit to study completion visit. The registry reports that the number of proteins that changed was zero.

For the endpoint calculation, protein exclusive intensity values, recorded in counts/second, were normalized using cyclic loess normalization. Mean abundance was determined for each protein, and mean Log2 Fold Change was calculated for each protein in the comparison between six weeks and baseline. Paired means were compared using a Limma t-test, and p-values were adjusted for multiple comparisons using the Benjamini-Hochberg method. The registry says 95% confidence intervals were calculated as a measure of dispersion, but does not list the interval estimates in the endpoint result.

In general, a protein-level endpoint is useful for investigating biological patterns and generating hypotheses about mechanisms. Counting proteins classified as changed provides a compact summary of the analysis, but it does not show the size or direction of each protein's change. Nor does the count alone indicate whether any observed change would be clinically important. The record does not state the precise decision threshold used to classify an individual protein as changed.

This endpoint is not a direct measure of diabetic complications or of a participant's clinical status. It addresses measured plasma protein abundance over the study interval. The result should therefore be interpreted within that laboratory and statistical scope.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous at the measurement level: protein abundance is derived from intensity values. The reported endpoint is a count of proteins classified as changed, but that count comes from statistical comparisons of abundance for individual proteins. In general, a continuous measurement retains information about magnitude, whereas a count of statistically classified measurements compresses that information. The registry reports the count but does not list each protein's abundance estimate or its change.

Why compare paired measurements?

The endpoint compares baseline and six-week measurements from the same participants, and the registry describes the analysis as a paired means comparison. In general, pairing links each later measurement to the corresponding baseline measurement for that participant. This can account for stable between-participant differences when estimating within-participant change. It does not remove time-related influences, measurement error, or other factors that may change between visits. Nor does pairing substitute for a separate control group.

The stated test is a Limma t-test. In general, a t-test evaluates whether the observed data are compatible with a null model of no mean difference, given the test's assumptions. The record identifies the test but does not provide the model details, assumptions checks, test statistics, or individual p-values. The registry also does not state the precise threshold used to call a protein changed.

What do normalization and fold change contribute?

Protein exclusive intensity values in counts/second were normalized using cyclic loess normalization. In general, normalization is intended to make measurements more comparable by addressing systematic differences in measurement distributions. A normalization step does not guarantee that every technical difference has been removed, and the record does not report diagnostic checks or other technical performance information.

The registry says mean protein abundance was determined from intensity values and that mean Log2 Fold Change was calculated for each protein between six weeks and baseline. In general, a fold change expresses a relative comparison between measurements, and a logarithmic scale can represent increases and decreases in a symmetric way around no change. The individual fold-change estimates are not included in the result description, so direction and magnitude cannot be evaluated from the reported count alone.

How should the confidence intervals be read?

The registry states that 95% confidence intervals were calculated as a measure of dispersion. In general, a confidence interval conveys the precision of an estimate under the method's assumptions; a narrower interval indicates greater precision than a wider interval. A 95% confidence interval does not mean that 95% of participants' values fall within it, and it does not establish the probability that the treatment caused a change. The actual intervals are not reported in the endpoint result description, so their widths and locations cannot be assessed.

Why adjust for multiple comparisons?

The analysis compared many individual proteins, and the registry says that p-values were adjusted using the Benjamini-Hochberg method. In general, when many hypotheses are tested, some results may appear statistically notable by chance even if there are no true differences. Benjamini-Hochberg adjustment is a method for controlling the expected proportion of false discoveries among results declared significant. The record does not say how many protein hypotheses were tested or specify the cutoff used to determine which proteins counted as changed. Those details matter for reproducing the classification.

For this endpoint, the registry's reported result is a count of zero proteins changed after the described analysis. The count should not be read as a direct estimate of average protein abundance, or as evidence that every baseline and endpoint measurement matched. The absence of protein-level estimates and adjusted p-values limits independent assessment of the analysis and of its precision.

Finally, the record notes that one participant sample lacked quantity to test. In general, missing measurements can affect precision and can introduce bias if missingness is related to participant characteristics or measured values. The record does not report why the sample was insufficient, how missingness was handled in the statistical procedure, or whether a sensitivity analysis was done. These omissions make it difficult to assess the robustness of the result beyond the stated analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry explicitly reports adjustment of p-values for multiple comparisons using the Benjamini-Hochberg method. This is relevant because the endpoint tests protein abundance across individual proteins rather than a single measurement. In general, multiplicity adjustment addresses the increased opportunity for false-positive findings when many hypotheses are tested. The record does not report the number of protein tests, the significance threshold, or the individual adjusted p-values, so the classification leading to the count of zero cannot be independently reconstructed from the listed result.

The endpoint population description says one participant sample lacked quantity to test. The record does not report further details about the reason for the insufficient quantity, how the missing sample was treated, whether a replacement or repeat assay was attempted, or whether the analysis used any imputation. It also does not report a sensitivity analysis addressing the missing sample. These are important distinctions because missing measurements can affect both the information available and the interpretation of the analyzed group.

The record does not provide a broader analysis-population definition, such as detailed inclusion or exclusion rules for the protein analysis. Nor does it report whether all measured proteins passed quality-control criteria or how such decisions were made. The only stated population-specific information for this endpoint is that one participant sample lacked quantity to test.

Although the endpoint description states that 95% confidence intervals were calculated, it does not list the intervals in the result description. The record also does not state the exact rule for identifying a protein as changed. As a result, the stated count can be reported, but the uncertainty and classification details for individual protein results remain unavailable in the registry record.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry lists serious adverse events for the fenofibrate group as 0 among 10 at risk. It also lists other adverse events as 0 among 10 at risk and deaths as 0. The record gives no serious adverse-event terms in its list of top serious terms. These figures describe what was recorded for this study; they do not establish that an event could not occur.

In general, adverse-event counts need to be read alongside the number of participants observed, the duration of observation, and the way events were collected and classified. This was a single-group study, with an endpoint timeframe of six weeks. The record does not provide further details on adverse-event definitions, collection procedures, severity grading, or event-specific follow-up. It also does not include a control group for comparison.

Zero recorded events is not the same as proof of zero risk. In a small study, uncommon events may not be observed, and the registry's figures do not provide a comparative safety estimate. The posted counts should therefore be understood as the recorded safety results for this study, not as a general estimate of the frequency of adverse events in other populations or settings.

12. What the Result Does, and Does Not, Mean

The primary result says that zero individual plasma proteins were reported as changed from baseline to the six-week endpoint under the registry's analysis. This is a finding about the protein-abundance endpoint and its statistical classification. It does not establish that every measured protein had exactly the same value at both visits. It also does not show the direction, magnitude, or precision of protein-level differences, because those individual results are not listed in the endpoint description.

The result does not establish that fenofibrate caused no biological change. The study has no separate control arm, and baseline comparison alone cannot distinguish the intervention from other changes over time or measurement variability. Conversely, the result does not establish that the intervention caused a beneficial change. The endpoint is not a clinical measure of diabetic retinopathy, kidney disease, nerve complications, or other health outcomes.

The registry's background describes a research interest in the potential mechanisms associated with fenofibrate and vulnerable tissues. That rationale is distinct from what the primary endpoint demonstrates. The reported result does not identify a mechanism, validate a biomarker, or show that any protein measurement predicts a clinical outcome. The record does not report such conclusions as outcomes of this analysis.

The study can be understood as a small, single-group investigation of within-participant plasma protein measurements over six weeks. It offers a descriptive result under the analysis stated in the registry. Missing sample quantity for one participant sample, unavailable protein-level estimates, unreported significance threshold, and absence of a control group all limit what can be inferred. The result is best kept within that scope: a reported count of zero proteins classified as changed, rather than a definitive conclusion about all biological effects or clinical consequences.

13. Why This Trial Matters Statistically

This study illustrates a common distinction in biostatistics: a biological measurement can be analyzed at many individual features, while the headline endpoint reports a single count of those features meeting a statistical rule. Here, the underlying measures are plasma protein intensity values, while the listed primary result is the number of proteins classified as changed. The count is concise, but it cannot substitute for individual estimates, uncertainty intervals, or the rule used to classify each protein.

It also illustrates why multiplicity matters in proteomic research. Testing many proteins creates an increased opportunity for chance findings, so the registry's use of Benjamini-Hochberg adjustment is an important part of interpreting the reported result. At the same time, without the number of tests, the adjusted p-values, or the decision threshold, readers cannot fully reproduce or audit the count.

The single-group paired design highlights another statistical distinction. Repeated measurements on participants can support an analysis of change over time, but they do not provide a concurrent control comparison. That matters when moving from an observed pattern to a causal claim about an intervention. The study therefore offers a useful example of how design determines the limits of interpretation, even when a statistical method for paired observations is specified.

Finally, the missing sample quantity and the lack of reported protein-level estimates show why transparent reporting of analysis populations and uncertainty is important. A zero count is interpretable as the posted result, yet further detail would be needed to judge precision, robustness, and reproducibility. These issues are central to evaluating biomarker and mechanism studies, where a statistical signal is only one step toward understanding clinical relevance.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The study's single-group design is the main limitation for causal interpretation. There is no separate control group, and the record lists allocation as NA and masking as none. A baseline-to-endpoint comparison can describe within-participant change but cannot establish how measurements would have changed over the same period without the intervention.

Enrollment was 10, and the endpoint population description notes that one participant sample lacked quantity to test. The registry does not explain why the sample was insufficient or how missingness was addressed. It does not report a sensitivity analysis, detailed analysis-population rules, or additional sample-quality information.

The posted primary result is a count rather than a set of protein-specific estimates. Although the endpoint description says 95% confidence intervals were calculated and mean Log2 Fold Change was calculated for each protein, it does not list those estimates or intervals. The record also does not state the threshold defining a changed protein, the number of protein tests, or individual adjusted p-values. This limits independent assessment of the reported zero count.

The registry does not report a sample-size justification, interim analyses, or a full statistical analysis plan. It does not provide enough information to assess whether the analysis decisions were prespecified or whether alternative analyses were considered. These details are not reported in the registry record.

The outcome is a laboratory protein-abundance measure over six weeks, not a direct clinical outcome. The registry's background discussion of diabetic complications supplies context for the research question but does not make those complications outcomes of this endpoint. The record therefore supports a narrow interpretation of the protein analysis, not conclusions about clinical benefit or long-term effects.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Fenofibrate in Type 2 Diabetes — NCT03829514.
• Registry record last updated 2023-10-24; status last verified 2023-10.
• Results first posted 2022-02-08.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT03829514, titled “Fenofibrate in Type 2 Diabetes,” is a completed Phase 4, single-group study sponsored by the Medical University of South Carolina. The record lists enrollment as 10, with one arm in which participants took fenofibrate. The study began on 2019-02-04, and primary completion was recorded as 2020-03-04.

The primary outcome was the number of individual plasma proteins that changed between baseline and the study-completion visit over six weeks. Protein abundance was measured by proteomic analysis using liquid chromatography-mass spectrometry, with paired means compared using a Limma t-test and p-values adjusted by the Benjamini-Hochberg method. The registry reports zero proteins changed. One participant sample lacked quantity to test.

The safety record lists 0 serious adverse events among 10 at risk, 0 other adverse events among 10 at risk, and 0 deaths. There was no separate control arm, and the reported endpoint concerns protein abundance rather than clinical complications. Protein-specific estimates, confidence interval values, the decision threshold for a changed protein, and a detailed missing-data analysis are not reported in the registry record.